The art of tracking motion of a subject in an activity with a camera is relatively well known. In general, a camera captures pluralities of discrete video images (frames) of the activity and each frame becomes analyzed.
With some motion tracking devices, a two-dimensional grid corresponding to the two-dimensions of the video image frames has data points plotted thereon that correspond to a particular feature of the subject. For example, to assess a horse stride length, it is important to know when each leg of the horse leaves the ground and when it returns to the ground. Thus, a horse's hoof as seen in each video frame may be plotted on the grid and a curve fit to the data points. Often times a transmitter device, which communicates to a receiver associated with a computing system to which the camera capturing the activity is attached, may be secured to the horse's hoof to assist in providing a cross-reference for the grid. Other times, reflectors or color contrasting devices attach to the horse's hoof.
This technique, however, suffers numerous shortcomings. For instance, each and every frame of the video must supply information about the horse's hoof to have an effective plot. This makes the technique labor intensive. Still further, the technique suffers because the subject, e.g., horse, may be running in a multi-horse race activity and in every frame the horse's hoof may not be fully visible. In turn, estimating the hoof position is impractical. Horses are also required to have the same equipment in a race and thus hoof-transmitters are not acceptable devices.
Accordingly, the art of motion tracking desires simple yet effective techniques for assessing actions of a subject in an activity while maintaining practicality.